I was alarmed because my best understanding of what I read is that there are two escape hatches in the covenant, should Microsoft care to make use of them. I wrote to Larry Rosen and asked him if he'd care to comment on what Andy wrote. I expressed that my deepest concern is about Phipps' number 4 and 6. Larry's comment and then Andy's reaction follow.

The bottom line is that my understanding was correct. As long as there are software patents, there is danger, if Microsoft were to choose to ... um.. . go bad, shall we say? The thing about patents that troubles me so is that openness as to a license is one thing, and very fine on its own, but if you are walking on a substratum of patents under your licensed feet, there is still a danger that you'll end up stepping into patent quicksand. More and more, I personally suspect that is the Microsoft strategy.

First, so you can understand Larry's comment, here are the two concerns from Phipps, abbreviated and italicized (so do visit Andy's blog for the full version), along with Andy's explanation in regular text, beginning with Phipps' concern #4 about Microsoft's covenant not to sue:

4. It is specific to the version currently existing, meaning I can be hooked into supporting it now, but when Office 12 or Office 13 comes out & I update to be compatible with the format in that I can get sued. The covenant Sun uses creates ongoing protection.

This is a serious one, and it's one that I've examined at length in two prior posts here and then here. First, in fairness to Microsoft's intention (although not its wordsmithing) and as I pointed out in the first of those posts, Microsoft does say at a separate page at its site that it will extend the same covenant to the Office 12 XML Reference Schemas. But that's as far as they've gone.

Again, to be fair, the way standards organizations work, someone can always drop out and not have to make a patent pledge regarding any new parts of a revision of a specification. But its undertakings are irrevocable and perpetual as to the version of the specification that was released while it was an involved member. ...

So one hopes that Ecma will do something to close this gap as a condition to accepting Microsoft's offer. . . .

Back to Simon one last time:

6. If the same form of words were used for a contribution to ECMA, then those prototyping the ongoing evolution of the standard as ECMA changed it would lose protection the instant any change was made. It applies only to Microsoft's input, not to ECMA's output. Or maybe they would rather ECMA didn't change anything?

...The fact of the matter is that Microsoft, so far, has only said on a Web page that it will make a covenant for the Office 12 versions of the XML Reference Schema -- and not for any Ecma standard based upon those Schema that may ultimately be adopted and issued. If it wishes, it can stop right where it is, and to the extent the Ecma working group adds any extensions, Microsoft could utilize the Ecma RAND-tolerant policy for those portions of the resulting Ecma standard....

But here's where the convergence of Simon's first and last question bring us to an interesting conclusion: if Ecma adds to the Microsoft contribution, then a "conformant" implementation of the Ecma standard would have a subset of itself that was "conformant" to the XML Reference Schema, and the implementation would be entitled to the covenant as to the conformant part, and RAND benefits as to the balance. But if instead of adding to, the Ecma working group changes part of the XML Reference Schema, then Microsoft could say "sorry – a critical exception has occurred" – and the covenant crashes.

If Andy is correct, I thought as I read this, then Microsoft has two holes big enough to drive a truck through. That's how I understood it, but not being a lawyer, I decided to ask Larry if I'd understood it correctly. Here is his response:

Hi PJ,

Thanks for the opportunity to comment on Andy Updegrove's latest blog [1] about the Microsoft covenant not to sue.

I can't quarrel with any of Andy's analysis. There are possible "escape hatches" for Microsoft in its covenant and in the Ecma process. The only thing that would guarantee otherwise would be if there were no software patents, but we're not at that point.

It doesn't surprise me that some people are suspicious of Microsoft and anticipate that company to pounce once it has trapped everyone into its web of patents. It earned that suspicion by past behavior although I'm told of encouraging changes there. I agree it is possible for Office 12 or Office 13 or some subsequent version to be outside the covenant and therefore once again proprietary. (Simon Phipps' point 4.) And certainly it is possible for the Ecma process to disintegrate into chaos and create a specification outside the covenant. (Simon's point 6.)

As a practical matter, I don't expect either alternative to occur. Microsoft has no reason to revert to locking up its XML document formats because doing so would only antagonize its own customers, not to speak of the State of Massachusetts and the open source community. Among the market requirements for Microsoft's subsequent versions of Office will be the availability of the Ecma-standard XML document format regardless of what add-on formats their software might offer. Another market requirement, I'm sure, will be support for the OASIS-standard XML document format. Customers of both Microsoft and Open Office products will demand the ability to exchange documents under both standards until, perhaps in some distant future, merger occurs on some new technology.

I now expect the Microsoft Office and Open Office products to compete on the merits without Microsoft and Sun suing people for practicing their XML document format patents. I expect that truce to continue indefinitely because it is in everyone's best interests.

Free software and open source software can forever implement certain Ecma-standard and OASIS-standard XML document formats. That's what I insisted upon in every conversation I had with Microsoft's attorneys and Sun's attorneys during all our past quarrels over patent license terms. I believe that's what we got from both companies.

That's why I'm pleased, even though I'm not naive enough to believe there aren't subsequent battles we'll have to fight about open standards. The best way to ensure ongoing cooperation is for open source companies to join the standardization efforts at Ecma and OASIS. Keep them open.

/Larry

[1] http://www.consortiuminfo.org/newsblog/blog.php?ID=1792

He can't quarrel with any of Andy's analysis, he writes. The escape hatches are real. Larry sent his reply to Andy also, and that made it possible for the conversation to continue. Here is Andy's reponse, again with his permission to share it with you:

Thanks for copying me, Larry.

So far, I've written three detailed blog entries from the "what could Microsoft do?" perspective, the latest being the one you note below. I haven't tried so far to guess what Microsoft will do, however, which is more the approach that you're taking with this email. That's just as valid an approach, but one for which I have less confidence in my abilities.

If I had to turn from analysis of possibilities to reading a crystal ball, however, I'd make the following additional prefatory observations:

Question one: Is Microsoft playing dirty?

1. I've talked to some of the top people at some of the major players involved, to lobbyists, and to reporters over the last three months, and I've asked many of them if they've run into direct proof of Microsoft being up to no good. So far, I haven't run into a single instance that someone actually knows about. Do they have lobbyists, publicists, trade associations and bloggers in their camp? Of course they do. And so do ODF advocates.

2. There are also people I know inside Microsoft who I know and like, and I don't think that they believe that they are engaged in an underhanded operation here.

3. With respect to Massachusetts, the local political situation is certainly breaking Microsoft's way. One wouldn't expect Microsoft not to take advantage of that if they could.

Net net, I don't have any reason to think that one side is playing markedly tougher ball than the other. If one is, then they are extremely good at covering their tracks. It's possible, of course, that Microsoft may be playing more skillfully, but if they're coloring inside the lines, then it's harder to complain. There's a lot at stake here, and business is business.

Question two: What will Microsoft do at Ecma?

1. Here I'm a bit less sanguine. Microsoft has a tightly integrated product that it's been developing for many, many years. I would not expect them to want to let anyone trifle with it.

2. In this context, the differences between the Sun covenant and the Microsoft covenant becomes more significant: Microsoft has covenanted not to sue those that implement its current schema, and has said that it will make the same covenant regarding the Office 12 schema. But it has explicitly *not* said that it will give a covenant regarding the final form of the schema that is approved by the Ecma membership.

3. In normal standard setting, it is usual for technology to be offered as the starting point, but not to give a blanket promise as to the entire final work product of the process. But under the rules of standard setting organization, it *is* required to irrevocably commit to patent terms as to so much of the material that it contributes that is included in the final specification. And the word *conformant* is not used in the submission - ever - in my personal experience, which spans scores of consortia.

Net net, when you add points 1 to 3 together, it is reasonable to assume that Microsoft is positioning itself to offer the Office 12 XML Reference Schema to Ecma for a rubber stamp of approval. If that would be as useful a result for the industry as a submission that could be pushed and pulled a bit, then there is really nothing for me to object to (or at least for me to object to seriously). I am not competent to make a judgment like that, but it would be very useful if someone else would.

It's very possible, of course that Microsoft is just taking one step at a time with its covenant. If that's the case, then all can be made clear in the document of submission that it makes to Ecma. It would be very useful if that would be made public.

At the end of the day, I agree with you that ultimately the market will decide as between two alternatives based upon the value that customers perceive in those alternatives, and not pay too much attention to any of the analysis that has come before in the blogs of the world. Having a huge installed base will certainly be an advantage in determining how that will play out, however, including with respect to how much effort ISVs put into creating software that would appeal to what will certainly be a much smaller ODF market for the indefinite future.

But again, you can't blame Microsoft for taking advantage of that dynamic, if they can make it work for them.

So here's my bottom line: I think that Microsoft will do what it needs to do to make its business strategy succeed, and I wouldn't expect them to do any more or less, with 40% of their profit at stake. But that's as clear as I can get my crystal ball to be.

Best regards,

Andy

Of course, lawyers are normally polite. It's a job requirement, to be able to deal with opponents without telling them what you think of them. Really, don't mock it. It's very, very important to the smooth running of the legal system. It's something I admire and believe in deeply, and I try to abide by the same principles.

Additionally, both attorneys have to deal with Microsoft, have done so in the past, and will do so in the future. I, on the other hand, have absolutely nothing to lose, so I'll tell you what I think this means, without any flowery bits. I don't mean to imply that they are not sincere, by the way. I know they are.

What I think it means is that OpenDocument Format is uniquely qualified to meet the Commonwealth of Massachusetts' stated policy as to what qualifies as an Open Standard for their purposes. You can hardly call a standard open, if it belongs to a single vendor, who can escape a covenant not to sue any time it wishes. Whether or not you believe that Microsoft is capable of such a dirty trick is immaterial. If it is possible, the "standard" can't match the openness of ODF, nor does it match what the Commonwealth has called for, as far as I can see.

I just wrote an article for LWN on this subject, "When Is a Standard Truly Open? -- Only When It's Universal," which is still subscription-only until next week, and the article begins like this:

What makes one standard open and not another? Massachusetts, when deciding to use the OpenDocument Format, as set forth in its Enterprise Information Technology Architecture (ETRM) document [PDF], set the bar here:

[Secretary of Administration & Finance for the Commonwealth of Massachusetts Eric] Kriss emphasized, however, that the state is not moving to open standards for economic reasons but to protect the right of the public to open and free access to public documents for the foreseeable future. "What we've backed away from at this point is the use of a proprietary standard and we want standards that are published and free of legal encumbrances, and we don’t want two standards," Kriss said.

... Is the Microsoft covenant not to sue, assuming it is someday offered for their new version of XML schemas, and their plans to submit their XML to standards bodies ECMA and ISO sufficient to meet the Massachusetts requirements for openness?... If not, will Massachusetts decide their bar was set too high, in order to include Microsoft? If the bar was set too high, which part shall we lop off?

The article then goes down the list to show that nothing in Kriss's list is expendable, in my opinion, and suggests that the real question isn't whether a standard is open so much as it's whether it is universal, whether everyone -- including the GPL -- can use it without fear or restriction. Massachusetts already has a definition of what it requires. According to my reading, Microsoft has not yet met it. Would they not have to close the two escape hatches, for starters, to qualify?

Larry relies on the marketplace to make that happen, but will Massachusetts politicians have the technical sophistication to realize the superiority of ODF? What have you observed? They might like to note that Gartner has just stated that ODF is superior to Microsoft's XML, by the way:

Gartner has suggested that while OpenDocument is the best XML office document standard out there, the announcement of rival 'open' standard from Microsoft could seriously hinder its growth...

However, Gartner recommends that enterprises looking for an open XML-based document format should adopt the OpenDocument format if they can exploit XML immediately because the ECMA specification is unlikely to appear for at least another year.

Andy also notes the superiority of Sun's covenant not to sue:

With respect to the current situation, however, the ODF standard is already in the marketplace, is already supported by several products, and has already been submitted to ISO for adoption. So if one were to put oneself in the position of a developer or a government agency, the analysis of my previous post becomes relevant: am I better off under the Sun covenant and the OASIS IPR policy, or am I better off (or at least no worse off) under the Microsoft covenant and pledge to seek Ecma adoption, and then ISO endorsement?

From this market perspective, Microsoft arguably needs to go farther than Sun, rather than less far, to provide equivalent comfort to those currently choosing between ODF and the Microsoft schemas. The reason is because Ecma has not yet taken control of the XML schemas, much less adopted them. Until then Microsoft still controls (and can therefore change) the formats, or could withdraw the formats from Ecma if it was not pleased with the direction that the Ecma working group wished to go.

Perhaps Tim Brays' suggestion to blend the two standards into one is the most practical:

The ideal outcome would be a common shared office-XML dialect for the basics—and it should be ODF (or a subset), since that’s been designed and debugged—then another extended vocabulary to support Microsoft features , whether they’re cool new whizzy features or mouldy old legacy features (XML Namespaces are designed to support exactly this kind of thing). That way, if you stayed with the basic stuff you’d never need to worry about software lock-in; the difference between portable and proprietary would be crystal-clear. And, for the basic stuff that everybody uses, there’d be only one set of tags.

This outcome is technically feasible. Who could possibly be against it?

Of course, if Microsoft is desiring to use patents to maintain their stranglehold on the market, then his suggestion will not appeal to them.

Can Microsoft fix this? Yes. Will it? Let's watch and see. I hope they do. But the fundamental question is always this with Microsoft, as you'll remember from the SenderID flap: can GNU/Linux and the GPL participate on an even playing field? If not, then does it deserve to be called a standard? And even if you answer yes, does it meet Massachusetts' stated requirements? Microsoft may close the two loopholes. But if Microsoft doesn't address these flaws, and if ECMA just rubberstamps their submission blindfolded, then I think we'd have to agree that it isn't open by any definition that we can understand.